6 digits seems perfectly fine for me. On Fri, 5 Oct 2018 at 14:26, Greg Landrum <greg.land...@gmail.com> wrote:
> > On Fri, Oct 5, 2018 at 2:42 PM Ivan Tubert-Brohman < > ivan.tubert-broh...@schrodinger.com> wrote: > >> In the newer "V3000", the atom line is not column-based, which I believe >> gives more freedom to implementers to decide the precision of the >> coordinates. You can force RDKit to write in this format by calling >> SetForceV3000(True) on your writer object. I tried it and I get 5 digits >> after the decimal point instead of 4, so at least that's a start. Looking >> at the RDKit code (function GetV3000MolFileAtomLine), it just writes the >> coordinates without setting the precision, so what you get is the default >> stringstream conversion. Here's where one could in principle adjust this >> precision, but there's clearly no API to do so at the moment. >> > > Yep. This is not currently possible without editing C++ code. > If there is a real use case for having more than 6 sig figs for atomic > positions (this is what is currently available), we can certainly come up > with a way to make it happen. I don't recall having seen any real-world > examples where that would be desirable. > >
_______________________________________________ Rdkit-discuss mailing list Rdkit-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
